def _storeOtherDependentAttrs(self):
GEDependentAttributesHolder._storeOtherDependentAttrs(self)
self._isSorted = self._geIter.isSorted()
self._hasCircularElements = self._geIter.hasCircularElements()
self._hasNoOverlappingElements = self._geIter.hasNoOverlappingElements()
self._hasUndirectedEdges = self._geIter.hasUndirectedEdges()
self._valDataType = self._geIter.getValDataType()
self._edgeWeightDataType = self._geIter.getEdgeWeightDataType()
self._geSource.__class__ = GtrackGenomeElementSource
def _storeOtherDependentAttrs(self):
GEDependentAttributesHolder._storeOtherDependentAttrs(self)
self._isSorted = self._geIter.isSorted()
self._hasCircularElements = self._geIter.hasCircularElements()
self._hasNoOverlappingElements = self._geIter.hasNoOverlappingElements(
)
self._hasUndirectedEdges = self._geIter.hasUndirectedEdges()
self._valDataType = self._geIter.getValDataType()
self._edgeWeightDataType = self._geIter.getEdgeWeightDataType()
self._geSource.__class__ = GtrackGenomeElementSource
def __init__(self, geSource):
try:
if not geSource.hasBoundingRegionTuples():
self._geSource = GEDependentAttributesHolder(geSource)
else:
self._geSource = geSource
except NotIteratedYetError:
self._geSource = geSource
try:
self._geSource.parseFirstDataLine()
self._emptyGeSource = False
except:
self._emptyGeSource = True
def _commonTestComposer(self, withTrackGESource, composerCls, suffix):
geSourceTest = self._commonSetup()
for caseName in geSourceTest.cases:
if not (caseName == suffix or \
(caseName.startswith(suffix) and caseName[len(suffix)] in ['_','.'])):
continue
if 'no_print' in caseName or \
withTrackGESource and ('no_track_extract' in caseName or \
caseName.endswith('_no_hb')):
print 'Test case skipped: ' + caseName
continue
print caseName
case = geSourceTest.cases[caseName]
testFn = self._writeTestFile(case)
sourceClass = case.sourceClass if case.sourceClass is not None else GenomeElementSource
genome = self.GENOME if withTrackGESource else case.genome
rawCaseGESource = sourceClass(testFn, genome, printWarnings=False)
caseGESource = GEDependentAttributesHolder(rawCaseGESource)
#actualSourceClass = caseGESource._geSource.__class__
if withTrackGESource:
for x in caseGESource:
pass
boundingRegionTuples = caseGESource.getBoundingRegionTuples()
boundingRegions = [br.region for br in boundingRegionTuples]
if boundingRegions == [] or all(br.chr is None
for br in boundingRegions):
boundingRegions = GlobalBinSource(self.GENOME)
trackName = self.TRACK_NAME_PREFIX + case.trackName
self._preProcess(trackName)
allowOverlaps = True if (
'start' in case.prefixList
) and not caseName.endswith('_compose_no_overlaps') else False
inputGESource = TrackGenomeElementSource(self.GENOME, trackName, boundingRegions, \
printWarnings=False, allowOverlaps=allowOverlaps)
def testSorting(self):
geSourceTest = self._commonSetup()
for caseName in geSourceTest.cases:
if not caseName.startswith('gtrack'):
continue
if 'no_sort' in caseName:
print 'Test case skipped: ' + caseName
continue
print caseName
print
case = geSourceTest.cases[caseName]
testFn = self._writeTestFile(case)
print open(testFn).read()
print
sortedContents = sortGtrackFileAndReturnContents(testFn, case.genome)
print sortedContents
sourceClass = GenomeElementSource if case.sourceClass is None else case.sourceClass
forPreProcessor = True if case.sourceClass is None else False
sortedGeSource = GEDependentAttributesHolder(sourceClass('sortedFile.gtrack', case.genome, \
forPreProcessor=forPreProcessor, \
printWarnings=False, \
strToUseInsteadOfFn=sortedContents))
reprIsDense = TrackFormat.createInstanceFromGeSource(sortedGeSource).reprIsDense()
if not reprIsDense:
self.assertEquals(sorted(case.assertElementList), [ge for ge in sortedGeSource])
else:
for ge in sortedGeSource:
pass
self.assertEquals(sorted(case.boundingRegionsAssertList), [br for br in sortedGeSource.getBoundingRegionTuples()])
inputGESource = TrackGenomeElementSource(self.GENOME, trackName, boundingRegions, \
printWarnings=False, allowOverlaps=allowOverlaps)
else:
inputGESource = rawCaseGESource
composer = composerCls(inputGESource)
contents = composer.returnComposed()
print contents
composedFile = NamedTemporaryFile('w', suffix='.' + suffix)
composedFile.write(contents)
composedFile.flush()
#print actualSourceClass.__name__
outputGESource = GEDependentAttributesHolder(
sourceClass(composedFile.name, genome, printWarnings=False))
if 'no_check_print' in caseName or withTrackGESource and 'no_check_track_extract' in caseName:
print 'No checks for case: ' + caseName
else:
caseGEs = [ge.getCopy() for ge in caseGESource]
outputGEs = [ge.getCopy() for ge in outputGESource]
isSortableGE = any(
getattr(caseGEs[0], x) is not None
for x in ['start', 'end']) if len(caseGEs) > 0 else False
if withTrackGESource and isSortableGE and not caseGESource.hasBoundingRegionTuples(
):
caseGEs = sorted(caseGEs)
self.assertGenomeElementLists(caseGEs, outputGEs)
self.assertListsOrDicts(
caseGESource.getBoundingRegionTuples(),
def _decorateGESource(self, geSource):
return GEDependentAttributesHolder(geSource)
class FileFormatComposer(object):
FILE_SUFFIXES = ['']
FILE_FORMAT_NAME = ''
def __init__(self, geSource):
try:
if not geSource.hasBoundingRegionTuples():
self._geSource = GEDependentAttributesHolder(geSource)
else:
self._geSource = geSource
except NotIteratedYetError:
self._geSource = geSource
try:
self._geSource.parseFirstDataLine()
self._emptyGeSource = False
except:
self._emptyGeSource = True
#self._emptyGeSource = True
#for ge in self._geSource:
# self._emptyGeSource = False
# break
@staticmethod
def matchesTrackFormat(trackFormat):
return MatchResult(match=False,
trackFormatName=trackFormat.getFormatName())
def composeToFile(self, fn, ignoreEmpty=False, **kwArgs):
ensurePathExists(fn)
f = open(fn, 'w')
ok = self._composeCommon(f, ignoreEmpty, **kwArgs)
f.close()
return ok
def returnComposed(self, ignoreEmpty=False, **kwArgs):
memFile = StringIO()
self._composeCommon(memFile, ignoreEmpty, **kwArgs)
return memFile.getvalue()
def _composeCommon(self, out, ignoreEmpty=False, **kwArgs):
if ignoreEmpty and self._emptyGeSource:
return False
self._compose(out, **kwArgs)
return True
def _compose(self, out, **kwArgs):
raise AbstractClassError()
def _commonFormatNumberVal(self, val):
if isNan(val) or val is None:
return '.'
return ('%#.' + str(OUTPUT_PRECISION) + 'g') % val
#return '%.5f' % val
def _commonFormatBinaryVal(self, val):
if val == BINARY_MISSING_VAL:
return '.'
return 1 if val == True else 0
@classmethod
def getDefaultFileNameSuffix(cls):
return cls.FILE_SUFFIXES[0]
def _commonTestComposer(self, withTrackGESource, composerCls, suffix):
geSourceTest = self._commonSetup()
for caseName in geSourceTest.cases:
if not (caseName == suffix or \
(caseName.startswith(suffix) and caseName[len(suffix)] in ['_','.'])):
continue
if 'no_print' in caseName or \
withTrackGESource and ('no_track_extract' in caseName or \
caseName.endswith('_no_hb')):
print 'Test case skipped: ' + caseName
continue
print caseName
case = geSourceTest.cases[caseName]
testFn = self._writeTestFile(case)
sourceClass = case.sourceClass if case.sourceClass is not None else GenomeElementSource
genome = self.GENOME if withTrackGESource else case.genome
rawCaseGESource = sourceClass(testFn, genome, printWarnings=False)
caseGESource = GEDependentAttributesHolder(rawCaseGESource)
#actualSourceClass = caseGESource._geSource.__class__
if withTrackGESource:
for x in caseGESource:
pass
boundingRegionTuples = caseGESource.getBoundingRegionTuples()
boundingRegions = [br.region for br in boundingRegionTuples]
if boundingRegions == [] or all(br.chr is None for br in boundingRegions):
boundingRegions = GlobalBinSource(self.GENOME)
trackName = self.TRACK_NAME_PREFIX + case.trackName
self._preProcess(trackName)
allowOverlaps = True if ('start' in case.prefixList) and not caseName.endswith('_compose_no_overlaps') else False
inputGESource = TrackGenomeElementSource(self.GENOME, trackName, boundingRegions, \
printWarnings=False, allowOverlaps=allowOverlaps)
else:
inputGESource = rawCaseGESource
composer = composerCls(inputGESource)
contents = composer.returnComposed()
print contents
composedFile = NamedTemporaryFile('w', suffix='.' + suffix)
composedFile.write(contents)
composedFile.flush()
#print actualSourceClass.__name__
outputGESource = GEDependentAttributesHolder(sourceClass(composedFile.name, genome, printWarnings=False))
if 'no_check_print' in caseName or withTrackGESource and 'no_check_track_extract' in caseName:
print 'No checks for case: ' + caseName
else:
caseGEs = [ge.getCopy() for ge in caseGESource]
outputGEs = [ge.getCopy() for ge in outputGESource]
isSortableGE = any(getattr(caseGEs[0], x) is not None for x in ['start','end']) if len(caseGEs) > 0 else False
if withTrackGESource and isSortableGE and not caseGESource.hasBoundingRegionTuples():
caseGEs = sorted(caseGEs)
self.assertGenomeElementLists(caseGEs, outputGEs)
self.assertListsOrDicts(caseGESource.getBoundingRegionTuples(), outputGESource.getBoundingRegionTuples())
if withTrackGESource:
self._removeAllTrackData(self.TRACK_NAME_PREFIX)
def testHeaderExpansion(self):
geSourceTest = self._commonSetup()
for caseName in geSourceTest.cases:
if not caseName.startswith('gtrack'):
continue
if 'no_expand' in caseName:
print 'Test case skipped: ' + caseName
continue
onlyGuaranteed = 'no_types_expanded' in caseName
print caseName
print '==========='
case = geSourceTest.cases[caseName]
headerLines = [line if not self._isHeaderLine(line) else
'##' + ': '.join([str(x).lower() for x in Gtrack.getHeaderKeyValue(line.strip())])
for line in case.headerLines]
fullContents = os.linesep.join(headerLines + case.lines)
print 'Original:\n\n' + fullContents
case.headerLines = [line for line in headerLines if not self._isExpandableHeader(line, onlyGuaranteed)]
print '-----'
print 'With headers removed:\n\n' + os.linesep.join(case.headerLines + case.lines)
testFn = self._writeTestFile(case)
expandedContents = expandHeadersOfGtrackFileAndReturnContents(testFn, case.genome, onlyNonDefault=False)
print '-----'
print 'With expanded headers:\n\n' + expandedContents
expandedContentsOnlyNonDefaults = expandHeadersOfGtrackFileAndReturnContents(testFn, case.genome, onlyNonDefault=True)
print '-----'
print 'With expanded headers (only non-default headers):\n\n' + expandedContentsOnlyNonDefaults
origExpandableHeaders = dict([Gtrack.getHeaderKeyValue(line) for line in headerLines \
if self._isExpandableHeader(line, onlyGuaranteed=False)])
notExpandableHeaders = dict([Gtrack.getHeaderKeyValue(line) for line in case.headerLines \
if self._isHeaderLine(line) and not self._isValueNotKeptHeader(line)])
expandedHeaders = dict([Gtrack.getHeaderKeyValue(line) for line in expandedContents.split(os.linesep) \
if self._isHeaderLine(line)])
if 'no_check_expand' in caseName:
print 'No checks for case: ' + caseName
else:
for header in origExpandableHeaders:
self.assertEquals(origExpandableHeaders[header], expandedHeaders[header])
for header in notExpandableHeaders:
self.assertEquals(notExpandableHeaders[header], expandedHeaders[header])
for contents in [expandedContents, expandedContentsOnlyNonDefaults]:
sourceClass = GenomeElementSource if case.sourceClass is None else case.sourceClass
forPreProcessor = True if case.sourceClass is None else False
stdGeSource = GEDependentAttributesHolder(sourceClass('expanded.gtrack', case.genome, \
forPreProcessor=forPreProcessor, \
printWarnings=False, \
strToUseInsteadOfFn=contents))
self.assertEquals(case.assertElementList, [ge for ge in stdGeSource])
self.assertEquals(case.boundingRegionsAssertList, [br for br in stdGeSource.getBoundingRegionTuples()])
def __init__(self, geSource, genome=None):
from gold.origdata.GEDependentAttributesHolder import GEDependentAttributesHolder
geSource = GEDependentAttributesHolder(geSource)
GESourceWrapper.__init__(self, geSource)
GenomeElementSource.__init__(self, '', genome=genome)